Google Data-Driven Attribution
Understand how Google Ads data-driven attribution assigns conversion credit, which conversion actions are eligible, how much data Google recommends, what the AnyTrack Conversion API feed sends to the model, and how the two click-based AnyTrack attribution models differ from it.
Data-driven attribution (DDA) is the default attribution model in Google Ads. The model decides how much credit each ad interaction receives for a conversion. It reads the conversion data that your Google Ads account received.
AnyTrack has no data-driven model. AnyTrack sends the conversions that the Google model reads.
What Data-Driven Attribution Is
Section titled “What Data-Driven Attribution Is”Data-driven attribution is a Google Ads attribution model. Google calculates the contribution of each ad interaction on the conversion path. To do this, Google compares the paths of customers who convert with the paths of customers who do not convert. The model then assigns fractional credit to each interaction.
A rule-based model applies a fixed rule instead. Last click gives all credit to the final ad interaction. Data-driven attribution applies no fixed rule, and the credit split changes when the underlying data changes.
Google applies the model to website conversions, store visits, and Google Analytics conversions. The supported networks are Search (including Shopping), YouTube, Display, and Demand Gen.
Eligibility and Data Volume
Section titled “Eligibility and Data Volume”Every conversion action is eligible. Google states: “All conversion actions are eligible for data-driven attribution (DDA), regardless of conversion or interaction volume.”
Volume changes the accuracy of the model, not the eligibility of the account.
| Question | Google position |
|---|---|
| Minimum volume to use the model | None. Every conversion action is eligible |
| Recommended volume | At least 200 conversions and 2,000 ad interactions in supported networks within a 30-day period |
| Below the recommendation | The model still functions. Google assigns credit with less precision |
What AnyTrack Sends to the Model
Section titled “What AnyTrack Sends to the Model”AnyTrack does not model attribution for Google Ads. AnyTrack sends the conversion data that the model reads. Conversion Mapping decides which events AnyTrack sends to Google Ads. For each mapped conversion, AnyTrack sends:
- The conversion event and its value.
- The Google click ID (
gclid,gbraid, orwbraid) that AnyTrack read from the landing page URL. - The hashed customer data that Enhanced Conversions uses.
Before AnyTrack sends a conversion, AnyTrack deduplicates it. One sender per platform prevents a double count on both the converting and the non-converting paths.
Why the Feed Changes What the Model Sees
Section titled “Why the Feed Changes What the Model Sees”The model reads the conversions that Google Ads received. Three gaps keep conversions out of that set. AnyTrack closes each one.
- Browser limits. If a browser blocks the Google Tag, the browser records no conversion. AnyTrack records the conversion on the server and sends it through the Conversion API.
- Conversions that happen outside the browser. A purchase confirmed in Shopify, a lead status change in HubSpot, and an affiliate network postback all happen on another server. AnyTrack receives each one as server-side tracking and sends it to Google Ads.
- Conversions with no active session. AnyTrack stores the click ID of the visit in the Data Store. When the conversion arrives days later, AnyTrack matches it to the original ad click and sends the conversion with its Google click ID.
Each closed gap adds a real conversion path to the data that the model compares. The model assigns credit on a more complete set of paths.
How AnyTrack Attribution Differs
Section titled “How AnyTrack Attribution Differs”AnyTrack reports with two click-based attribution models: First Ad Click and Last Ad Click. Switch between them with the Attribution dropdown in the Campaign Report. AnyTrack has no data-driven model, no multi-touch model, and no post-view model.
The two systems answer different questions. Read them together, not as one number.
| Google Ads | AnyTrack | |
|---|---|---|
| Attribution model | Data-driven attribution by default. Rule-based models are also available | First Ad Click or Last Ad Click |
| Credit split | Fractional credit across the ad interactions on the path | Full credit to one ad click |
| Post-view conversions | Counted | Not counted |
| Conversion modeling | Google estimates conversions that it could not observe | None. AnyTrack reports the conversions it received |
| Where you read the result | The Conversions column in Google Ads, and Attributed Results in the Campaign Report | The conversion counts in the Campaign Report |
The two numbers differ for these reasons. Expect a gap between them. For the full list of causes, read Google Ads Conversion Goals.
Next Steps
Section titled “Next Steps”Send hashed customer data with each mapped conversion so Google matches more of them.
Data-Driven Attribution FAQ
FAQ was last reviewed on 2026-09-29